Plasma free fatty acids (non-esterified fatty acids) increase in the first hour of the onset of acute myocardial ischaemia. This results from catecholamine stimulation of adipose tissue lipolysis. It can lead to a metabolic crisis in the injured myocardium with the development of ventricular arrhythmias and increased early mortality. Preconditioning, β-adrenergic blockade and glucose–insulin–potassium are possible therapeutic approaches, but anti-lipolytic agents, such as some nicotinic acid derivatives, can reduce plasma free fatty acid concentrations within minutes and have untried potential. A clinical trial of their effectiveness is needed from the first moment when a patient with an acute coronary syndrome is seen by paramedics.
Many deaths, which occur during acute myocardial ischaemia as infarction develops, need not take place. Tremendous headway has been made by interventionist cardiologists with coronary angiography, insertion of stents, bypass surgery and the early infusion of thrombolytic agents–all with the aim of improving blood supply. But there have been few interventions directed specifically at the injured myocardium. Yet, it is a functioning myocardium that will determine survival .
Its viability depends on two factors. These are the revival of adequate coronary blood by measures described above and the restoration and maintenance of normal myocardial metabolism in the face of acute reduction in available oxygen. It is the latter which is missing from our therapeutic attack and the subject of the present review. Profound changes occur in systemic metabolism during any episode of acute ischaemia and these may result in the myocardium no longer receiving the optimum balance of energy substrates which will allow it to contract and function normally. This may lead to lethal ventricular fibrillation. To put it simply, pumps will not work if they are deprived of appropriate energy.
Normally, efficient aerobic myocardial metabolism depends on the availability of the key substrates, free fatty acids (FFA or non-esterified fatty acids) and glucose. FFA are responsible for 60%–70% of ATP produced and glucose for about 20%–25%, with lactate and ketones making smaller contributions. The rate of ATP breakdown is balanced by that of ATP synthesis. This cycle depends on the efficiency of myocardial oxygen consumption. This is easily met in aerobic conditions, but less so or not at all when oxygen supply is reduced. Reduced coronary flow leads to myocardial ischaemia with local impairment of ATP formation. Ischaemia occurs regionally according to the extent of coronary arteriole underperfusion and local reperfusion. The areas affected may be quite small and probably change rapidly, since myocardial energy kinetics are in continuous flux . But they can be the site of depolarization and arrhythmias.
The availability and utilization of substrates locally is unpredictable . The fatty acid uptake by the myocardium can determine myocardial oxygen requirements . An excess of FFA in some underperfused areas could be temporarily deleterious and increase local myocardial oxygen consumption patchily, leading to the development of gradients in substrate utilization and electrolyte transfer, with temporary interruption of the distribution of action potential as well as impairment of contractility [5,6]. β-Oxidation may be impaired, possibly irreversibly . Increased FFA also suppress glucose oxidation through inhibition of pyruvate dehydrogenase and, in severely hypoxic localized areas of the myocardium, impaired glucose utilization and uptake, due to insulin suppression, may worsen the oxygen wasting effects of increases in myocardial FFA concentrations .
FREE FATTY ACIDS
Plasma FFA concentrations are directly related to its release from adipose tissue . The turnover is exceedingly rapid with a half-time of 2 min. Circulating FFA are bound with various degrees of affinity to albumin. Saturation of the two principal binding sites occurs at approximately 1.2 mmol/l. There is an exponential tissue uptake of FFA above this level and, when higher molar ratios are reached, the uptake by all tissues  including ischaemic areas of the myocardium is increased, with a greater risk of ventricular fibrillation. In isolated rat hearts, FFA can be directly arrhythmogenic if the molar ratio to albumin is high even in the absence of ischaemia . The importance of the molar relationship between FFA and albumin is well illustrated by the demonstration that lipid-free albumin infusions that reduce the FFA/albumin ratio simultaneously decrease the extent of ST elevation in dogs with coronary occlusion .
In 1963, when studying electrophoretic analyses of lipoproteins, we identified that an increase in albumin-bound FFA occurred commonly in patients with acute myocardial infarction. Subsequently, we reported that an increase in plasma FFA also occurred in shock, renal colic, non-myocardial causes of pain and in cerebral infarction . We regarded the rise in plasma FFA as an ubiquitous non-specific response to stress–a catecholamine-induced response.
During any acute coronary syndrome, a moderate increase in catecholamine activity will occur and augment inotropy and help to maintain contractility in the face of impaired myocardial oxygenation. But acute stress, fear (angor animi) and pain of a developing coronary syndrome may produce excess catecholamine activity and can lead to an unusually high FFA/albumin ratio with excess uptake of FFA by the hypoxic myocardium. Plasma adrenaline concentrations increase within minutes of the onset of an acute coronary syndrome  and remain elevated for up to 20 h, depending on the severity of the response to stress. This stimulates tissue lipolysis of stored glycerides and increases circulating FFA concentrations. It is not surprising, therefore, the plasma FFA rise very rapidly and can double or treble resting values during the first hours of acute myocardial infarction. Fatty acid oxidation uses more oxygen per mole than glucose. In the liver, glycogenolysis is increased but pancreatic insulin secretion is reduced and a relative glucose debt can occur locally. Additionally adrenaline has a direct inhibitory effect on glucose-mediated secretion of insulin from β-cells. Plasma cortisol and cyclic AMP concentrations also rise rapidly .
ELEVATED PLASMA FFA AND DEATH
In 1968, we studied 200 consecutive patients with acute myocardial ischaemia and reported that those who had particularly high plasma concentrations of FFA also had an increased incidence of ventricular arrhythmias and a higher early mortality (Figure 1) . This association was soon confirmed . The increase in plasma FFA occurs within an hour after the onset of symptoms of acute myocardial ischaemia (Figure 2) . It does not result from existing myocardial damage as with the efflux of troponins and transaminases but is a highly sensitive biomarker of developing injury. In my opinion, the measurement of FFA concentrations should be made the first time when blood is taken from a patient with suspected acute myocardial ischaemia since it is an important immediate indicator of immediate prognosis. Kit methods are now available for rapid measurement of FFA concentrations.
Maximum serum FFA concentrations in 200 consecutive men with acute myocardial infarction (M.I.) and age-matched controls
The increase in plasma FFA occurs within 1 h after the onset of symptoms of acute myocardial ischaemia and precedes those of cardiac markers that result from myocardial damage
Indirect confirmation of the relation between high plasma FFA and mortality has come from the Paris Prospective Study of 5250 men . After 22 years of follow-up, an increase in circulating FFA at baseline examination was significantly related to subsequent sudden death, defined as natural death occurring within 1 h after the onset of acute symptoms. The authors regarded this correlation as a manifestation of increased adrenergic tone. Furthermore, as we showed for ventricular fibrillation, the risk of sudden death increased with increasing values of FFA. There was no correlation in the Paris Study between plasma FFA concentrations and causes of fatal myocardial infarction other than that of sudden death. Another large study showed that, over a follow-up time of 6.85 years, elevated FFA predicted sudden cardiac death in 3315 patients who had had coronary angiography . But this correlation did not apply in 4657 men and women over the age of 75 from the Cardiovascular Health Study  possibly because they were a self-limiting population.
Recently, data from the TIMI II trial of thrombolysis for STEMI, have shown this is also true for unbound FFA (FFAu) . FFAu is a very small fraction of total FFA. It increases exponentially in parallel with increasing total albumin-bound FFA and is more sensitive to physiological changes. In 1834 patients, the upper two quartiles of FFAu were associated with a significantly higher mortality at 1 day, 7 and 30 days after acute myocardial infarction. In this study, high FFAu at baseline was a risk factor independent of age, gender, race, body mass index (BMI), diastolic and systolic blood pressure, history of diabetes, previous MI and hypertension and use of β-blocking drugs. Even more recently, data from the Oregon Sudden Unexpected Death Study also show that elevated plasma FFA levels predicted sudden death .
The precise mechanism through which excessive myocardial uptake of plasma FFA might lead to ventricular fibrillation is speculative though there is compelling evidence to suggest that the rise in FFA is the primary cause. Perfusion of isolated rat hearts with high-molar ratios of albumin-bound fatty acids has a direct arrhythmogenic effect . Intravenous injection of long-chain saturated fatty acids into anaesthetized healthy dogs induced ventricular arrhythmias . Infusions of lipid-free albumin to dogs with coronary artery occlusion resulted in rapid reduction in myocardial uptake of FFA and of ST segment elevation in contrast to the effects of infusion of normal albumin . Heparin-induced plasma lipolysis can also lead to ventricular fibrillation in dogs and heparin-induced ventricular arrhythmias can be prevented and reversed by protamine sulfate . Heparin activates chylomicron and triacylglycerol (triglyceride) lipolysis and, in the presence of postprandial hypertriglyceridaemia, leads to high plasma FFA concentrations. Hypothetically, the use of heparin in the management of an acute coronary syndrome might, if there is concurrent postprandial lipidaemia, favour the development of ventricular fibrillation.
FATTY ACID TOXICITY
The mechanisms of fatty acid toxicity are complex. When there is hypoxia, free radicals may occur with a membrane detergent effect [7,8,25]. A regional excess of fatty acids may lead locally to peroxidation of membranes with dispersion of membrane potentials, and activation of cytokines. Plasma FFA enter cardiomyocytes and thence into mitochondria where they may uncouple mitochondrial respiration [7,25]. During ischaemia, β-oxidation of lipids in mitochondria may also be inhibited with accumulation of long-chain acylcarnitine and acyl-CoA . The accumulation of detergent CoA derivatives and lysophospholipids resulting from instability of membrane lipids also favours the development of arrhythmias. FFA may inhibit the Na+, K+ ATPase pump, leading to high intracellular Na+ and Ca2+ . This could lead to cytosolic Ca2+ overload with the occurrence of electrical re-entry and arrhythmias. Excess FFA may lead to accumulation of extracellular K+ and shortening of action potential. Additionally, the activity of the insulin-responsive glucose transporter (GLUT4) falls in the presence of excess FFA . Thus, elevated FFA and intracellular lipid reduce insulin-stimulated glucose transport. High FFA levels also impair capillary recruitment and acetylcholine-mediated vasodilatation .
Not all fatty acids behave similarly, and not all are pro-arrhythmic. Some polyunsaturated fatty acids have an anti-arrhythmic action. There have been consistent observations that n−3 polyunsaturated fatty acids from fish oils (particularly eicosapentaenoic and docosahexaenoic acids) decrease the tendency of experimentally induced myocardial ischaemia to develop ventricular fibrillation [30,31] and the GISSI study showed that they reduce the incidence of sudden cardiac death . The blood levels of n−3 fatty acids taken at baseline were found in a 17-year follow-up of the Physicians Health Study to be inversely related to sudden cardiac death . The anti-arrhythmic effects of n−3 fatty acids may be explained by their effect on several basic electrophysiological mechanisms . They can lead to Na+ channel inhibition and prolongation of refractory periods in cardiomyocytes.
This might interfere with re-entry circuits. Polyunsaturated fatty acids, not exclusively n−3, appear to act through stabilizing cardiac myocytes by modulating conductance of ion channels in the sarcolemma, particularly the fast, voltage-dependent Na+ current and the L-type Ca2+ currents, though other ion currents are also affected. All prostaglandins and thromboxanes produced from arachidonic acid (n−6) were found to be potent arrhythmogenic agents, whereas none of the comparable 3-series cyclo-oxygenase products of eicosapentaenoic acid were arrhythmogenic .
HOW BEST SHOULD THE EARLY METABOLIC CRISIS BE MINIMIZED?
The aim should be to reduce elevated FFA concentrations and to restore normal myocardial metabolism as quickly as possible and within an hour or two of the onset of acute ischaemia. There are several different approaches (Figure 3).
Flow diagram indicating therapeutic options to control high FFA concentrations and improve ischaemic myocardial metabolism
Although preconditioning is likely to improve local collateral blood supply, there is no evidence that it corrects the myocardial imbalance of substrates which occurs in the first hours of ischaemia. Though remote ischaemic preconditioning during evolving ST-elevation myocardial infarction improves myocardial salvage at 30 days after primary percutaneous coronary intervention, there was no difference in early deaths or reinfarction .
β-Adrenergic blockade might seem a logical approach to limit any adrenaline hyper-reaction, but, during developing myocardial infarction, it may lead to reduction in myocardial inotropy and to hypotension. Further, plasma FFA concentrations are only slowly reduced by β-adrenergic blockade. There have not been any trials specifically focused on its effect on the incidence of ventricular arrhythmias or on death during the first acute 6 h but we know from the COMMIT trial of 45852 patients that metropolol does not reduce sudden cardiac death or ventricular fibrillation occurring within the first 24 h but increases cardiogenic shock at that time .
Far more focus is needed on redressing the imbalance in myocardial metabolism that occurs during acute ischaemia . Direct strategies to increase glucose availability and uptake, or reduce adrenaline-stimulated tissue lipolysis and plasma FFA concentrations, are the best bet  and based on sound physiological principles . There have been many trials of the effectiveness of glucose–insulin–potassium (GIK) with mixed results. The first results of the biggest (IMMEDIATE) trial, which was stopped prematurely, are disappointing . Among patients with suspected acute coronary syndrome, out-of-hospital administration of intravenous GIK, compared with placebo, did not reduce progression to myocardial infarction. Also, GIK administration was not associated with improvement in 30-day survival but was associated with lower rates of the composite outcome of cardiac arrest or in-hospital mortality. These results were possibly due to insufficient numbers which made it underpowered for measurement of the primary endpoints. Also, the GIK infusions were initiated too late. For GIK to be effective, the infusion should be started as early as possible, preferably within the first 2–3 h. GIK infusions not only increase glucose availability but also reduce circulating FFA by inhibiting its release from adipose tissue, as found in the IMMEDIATE trial. However, the one-year outcomes look more promising . In patients with ST elevation myocardial infarction, the composites of cardiac arrest or 1-year mortality, and of cardiac arrest, mortality, or heart failure hospitalization within 1 year, were significantly reduced. These support the hypothesis that restoration of impaired myocardial metabolism is literally vital when acute ischaemia co-exists. The possible benefits of giving insulin alone in the earliest hours has not been explored.
The focus now should be more directly on metabolic measures that maintain myocardial ATP levels and availability, particularly since there appear to be few opportunities for benefit by additional reductions in door-to-balloon times . The lipolytic action of catecholamines on adipose tissue glycerides leading to efflux of FFA has long been known . Since it is undesirable to block catecholamines at a time when the acutely ischaemic myocardium needs their inotropic action, the key question is whether tissue lipolysis can be controlled sufficiently to produce rapid reductions in plasma concentrations of FFA and FFAu. If lipolysis can be inhibited early during an acute coronary syndrome, it is predictable that fewer ventricular arrhythmias will occur and that the earliest mortality will be reduced.
The lipase-inhibiting effect of nicotinic acid has been largely forgotten. More than 50 years ago, Carlson and Oro  demonstrated that nicotinic acid led to a ‘metabolic type of sympathicolysis’ with a rapid decrease in FFA. This was soon confirmed  and we have shown that arrhythmias and ST segment elevation or depression occurring during myocardial ischaemia can be reduced by a nicotinic acid analogue [44,45]. Nicotinic acid also depresses FFA extraction across the ischaemic myocardium during basal and isoprenaline stimulated lipolysis . These issues have been brought into focus by the proposal that there are nicotinic acid receptors which can be modulated [47,48].
Nicotinic acid and nicotinuric acid, but not nicotinamide , inhibit tissue lipoprotein lipases reducing triacylglycerol hydrolysis, probably through activation of specific Gi-coupled nicotinic acid receptors . This action of nicotinic acid is very rapid and results in immediate and profound inhibition of FFA efflux from adipose tissue. It can also lead to interruption of the myocardial energy-wasting glyceride cycle. A single oral dose of 1 g of nicotinic acid reduces plasma FFA levels within 3 min and for 3 h . It is, therefore, a potentially potent tool for the very early management of acute myocardial ischaemia.
The recent null result of the use of niacin over 3.9 years to prevent cardiovascular disease and myocardial infarction in a randomized trial of 25673 high-risk patients (HPS2-THRIVE)  is not relevant to the argument proposed here. My proposal is that the immediate use of nicotinic acid or a derivative will prevent ventricular arrhythmias occurring within the first hours of an acute coronary syndrome. The action of niacin, as used, is too slow to expect any benefit. Nor does the negative result of that trial undermine the strategy of this proposal since niacin (with laropripant to reduce vasodilation) was not initiated during this crucial time period.
Since nicotinic acid also causes hypotension, flushing and vasodilatation which can have adverse effects on the haemodynamic requirements during an acute myocardial infarct, a priority now is the synthesis of analogues without these undesirable vasomotor actions and, later, their clinical testing. We already know that 5-fluoro-3-hydroxymethylpyridine hydrochloride  and β-pyridyl-carbinol  are rapidly effective in reducing plasma FFA levels and ST elevation. The aim now should be the development of effective nicotinic acid analogues or receptor agonists retaining antilipolytic activity. A further potential problem is the rebound of lipolysis as nicotinic acid treatment continues but, in our experience, this is delayed for 15–20 h , in other words, well after arrival in the catheter laboratory. This can be dealt with in hospital and I do not recommend a slow release compound. It is important to try to control elevated FFA concentrations as early as possible. Another potential approach to the problem of excessive FFA is the possibility that fatty acid transport across cell membranes might be inhibited.
The design of an exploratory clinical trial of the effectiveness of a nicotinic acid analogue in the very early treatment of acute coronary syndromes need not be complex. Since oral nicotinic acid reduces plasma FFA so rapidly, administration of active analogues needs only to cover the time of the highest incidence of ventricular arrhythmias, i.e. from the arrival of paramedics up to coronary angiography or percutaneous catheterization–usually less than 6 h.
There is a strong case for reducing the risk of early arrhythmic death during acute myocardial ischaemia by controlling raised plasma FFA concentrations. This is probably best achieved through drugs that will rapidly inhibit tissue lipolysis without haemodynamic side effects. The nicotinic acid family is the obvious choice for pharmaceutical companies to take up this challenge.